End of training
Browse files- README.md +6 -4
- all_results.json +15 -0
- eval_results.json +9 -0
- runs/May19_22-08-09_tz579-raptorlake/events.out.tfevents.1716177937.tz579-raptorlake.65634.1 +3 -0
- train_results.json +9 -0
- trainer_state.json +1873 -0
README.md
CHANGED
@@ -2,6 +2,8 @@
|
|
2 |
license: apache-2.0
|
3 |
base_model: facebook/wav2vec2-base
|
4 |
tags:
|
|
|
|
|
5 |
- generated_from_trainer
|
6 |
datasets:
|
7 |
- timit_asr
|
@@ -14,11 +16,11 @@ model-index:
|
|
14 |
name: Automatic Speech Recognition
|
15 |
type: automatic-speech-recognition
|
16 |
dataset:
|
17 |
-
name:
|
18 |
type: timit_asr
|
19 |
config: clean
|
20 |
split: test
|
21 |
-
args:
|
22 |
metrics:
|
23 |
- name: Wer
|
24 |
type: wer
|
@@ -30,9 +32,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
30 |
|
31 |
# wav2vec2-base-timit-fine-tuned
|
32 |
|
33 |
-
This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the
|
34 |
It achieves the following results on the evaluation set:
|
35 |
-
- Loss: 0.
|
36 |
- Wer: 0.4173
|
37 |
|
38 |
## Model description
|
|
|
2 |
license: apache-2.0
|
3 |
base_model: facebook/wav2vec2-base
|
4 |
tags:
|
5 |
+
- automatic-speech-recognition
|
6 |
+
- timit_asr
|
7 |
- generated_from_trainer
|
8 |
datasets:
|
9 |
- timit_asr
|
|
|
16 |
name: Automatic Speech Recognition
|
17 |
type: automatic-speech-recognition
|
18 |
dataset:
|
19 |
+
name: TIMIT_ASR - NA
|
20 |
type: timit_asr
|
21 |
config: clean
|
22 |
split: test
|
23 |
+
args: 'Config: na, Training split: train, Eval split: test'
|
24 |
metrics:
|
25 |
- name: Wer
|
26 |
type: wer
|
|
|
32 |
|
33 |
# wav2vec2-base-timit-fine-tuned
|
34 |
|
35 |
+
This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the TIMIT_ASR - NA dataset.
|
36 |
It achieves the following results on the evaluation set:
|
37 |
+
- Loss: 0.4275
|
38 |
- Wer: 0.4173
|
39 |
|
40 |
## Model description
|
all_results.json
ADDED
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 20.0,
|
3 |
+
"eval_loss": 0.42749759554862976,
|
4 |
+
"eval_runtime": 39.6053,
|
5 |
+
"eval_samples": 1344,
|
6 |
+
"eval_samples_per_second": 33.935,
|
7 |
+
"eval_steps_per_second": 33.935,
|
8 |
+
"eval_wer": 0.41728125284530637,
|
9 |
+
"total_flos": 2.1476719263248095e+18,
|
10 |
+
"train_loss": 0.8618391515622879,
|
11 |
+
"train_runtime": 3159.4128,
|
12 |
+
"train_samples": 3696,
|
13 |
+
"train_samples_per_second": 23.397,
|
14 |
+
"train_steps_per_second": 0.734
|
15 |
+
}
|
eval_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 20.0,
|
3 |
+
"eval_loss": 0.42749759554862976,
|
4 |
+
"eval_runtime": 39.6053,
|
5 |
+
"eval_samples": 1344,
|
6 |
+
"eval_samples_per_second": 33.935,
|
7 |
+
"eval_steps_per_second": 33.935,
|
8 |
+
"eval_wer": 0.41728125284530637
|
9 |
+
}
|
runs/May19_22-08-09_tz579-raptorlake/events.out.tfevents.1716177937.tz579-raptorlake.65634.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:761f8f6656c0c227f5c72fd2abed63841c5757356b4cb775dfa24da593234fff
|
3 |
+
size 406
|
train_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 20.0,
|
3 |
+
"total_flos": 2.1476719263248095e+18,
|
4 |
+
"train_loss": 0.8618391515622879,
|
5 |
+
"train_runtime": 3159.4128,
|
6 |
+
"train_samples": 3696,
|
7 |
+
"train_samples_per_second": 23.397,
|
8 |
+
"train_steps_per_second": 0.734
|
9 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,1873 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 20.0,
|
5 |
+
"eval_steps": 100,
|
6 |
+
"global_step": 2320,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.08620689655172414,
|
13 |
+
"grad_norm": 9.595185279846191,
|
14 |
+
"learning_rate": 9e-07,
|
15 |
+
"loss": 9.1142,
|
16 |
+
"step": 10
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.1724137931034483,
|
20 |
+
"grad_norm": 9.732986450195312,
|
21 |
+
"learning_rate": 1.9e-06,
|
22 |
+
"loss": 8.3446,
|
23 |
+
"step": 20
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.25862068965517243,
|
27 |
+
"grad_norm": 14.272214889526367,
|
28 |
+
"learning_rate": 2.8000000000000003e-06,
|
29 |
+
"loss": 8.6592,
|
30 |
+
"step": 30
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.3448275862068966,
|
34 |
+
"grad_norm": 15.0160493850708,
|
35 |
+
"learning_rate": 3.8e-06,
|
36 |
+
"loss": 7.6985,
|
37 |
+
"step": 40
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.43103448275862066,
|
41 |
+
"grad_norm": 16.610979080200195,
|
42 |
+
"learning_rate": 4.800000000000001e-06,
|
43 |
+
"loss": 6.9688,
|
44 |
+
"step": 50
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.5172413793103449,
|
48 |
+
"grad_norm": 17.26924705505371,
|
49 |
+
"learning_rate": 5.8e-06,
|
50 |
+
"loss": 6.232,
|
51 |
+
"step": 60
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.603448275862069,
|
55 |
+
"grad_norm": 11.347734451293945,
|
56 |
+
"learning_rate": 6.800000000000001e-06,
|
57 |
+
"loss": 4.7271,
|
58 |
+
"step": 70
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.6896551724137931,
|
62 |
+
"grad_norm": 4.237112045288086,
|
63 |
+
"learning_rate": 7.8e-06,
|
64 |
+
"loss": 3.7919,
|
65 |
+
"step": 80
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.7758620689655172,
|
69 |
+
"grad_norm": 1.8833028078079224,
|
70 |
+
"learning_rate": 8.8e-06,
|
71 |
+
"loss": 3.3967,
|
72 |
+
"step": 90
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.8620689655172413,
|
76 |
+
"grad_norm": 1.3788093328475952,
|
77 |
+
"learning_rate": 9.800000000000001e-06,
|
78 |
+
"loss": 3.1618,
|
79 |
+
"step": 100
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.8620689655172413,
|
83 |
+
"eval_loss": 3.1117007732391357,
|
84 |
+
"eval_runtime": 40.0512,
|
85 |
+
"eval_samples_per_second": 33.557,
|
86 |
+
"eval_steps_per_second": 33.557,
|
87 |
+
"eval_wer": 1.0,
|
88 |
+
"step": 100
|
89 |
+
},
|
90 |
+
{
|
91 |
+
"epoch": 0.9482758620689655,
|
92 |
+
"grad_norm": 1.729278802871704,
|
93 |
+
"learning_rate": 1.08e-05,
|
94 |
+
"loss": 3.0865,
|
95 |
+
"step": 110
|
96 |
+
},
|
97 |
+
{
|
98 |
+
"epoch": 1.0344827586206897,
|
99 |
+
"grad_norm": 1.905969500541687,
|
100 |
+
"learning_rate": 1.18e-05,
|
101 |
+
"loss": 3.0809,
|
102 |
+
"step": 120
|
103 |
+
},
|
104 |
+
{
|
105 |
+
"epoch": 1.1206896551724137,
|
106 |
+
"grad_norm": 0.8360918760299683,
|
107 |
+
"learning_rate": 1.2800000000000001e-05,
|
108 |
+
"loss": 3.0346,
|
109 |
+
"step": 130
|
110 |
+
},
|
111 |
+
{
|
112 |
+
"epoch": 1.206896551724138,
|
113 |
+
"grad_norm": 0.7653716206550598,
|
114 |
+
"learning_rate": 1.3800000000000002e-05,
|
115 |
+
"loss": 3.0106,
|
116 |
+
"step": 140
|
117 |
+
},
|
118 |
+
{
|
119 |
+
"epoch": 1.293103448275862,
|
120 |
+
"grad_norm": 0.94779372215271,
|
121 |
+
"learning_rate": 1.48e-05,
|
122 |
+
"loss": 3.0165,
|
123 |
+
"step": 150
|
124 |
+
},
|
125 |
+
{
|
126 |
+
"epoch": 1.3793103448275863,
|
127 |
+
"grad_norm": 0.8457741737365723,
|
128 |
+
"learning_rate": 1.58e-05,
|
129 |
+
"loss": 3.0,
|
130 |
+
"step": 160
|
131 |
+
},
|
132 |
+
{
|
133 |
+
"epoch": 1.4655172413793103,
|
134 |
+
"grad_norm": 1.4369837045669556,
|
135 |
+
"learning_rate": 1.6800000000000002e-05,
|
136 |
+
"loss": 2.9903,
|
137 |
+
"step": 170
|
138 |
+
},
|
139 |
+
{
|
140 |
+
"epoch": 1.5517241379310345,
|
141 |
+
"grad_norm": 1.8290436267852783,
|
142 |
+
"learning_rate": 1.78e-05,
|
143 |
+
"loss": 2.9852,
|
144 |
+
"step": 180
|
145 |
+
},
|
146 |
+
{
|
147 |
+
"epoch": 1.6379310344827587,
|
148 |
+
"grad_norm": 1.1530190706253052,
|
149 |
+
"learning_rate": 1.88e-05,
|
150 |
+
"loss": 2.99,
|
151 |
+
"step": 190
|
152 |
+
},
|
153 |
+
{
|
154 |
+
"epoch": 1.7241379310344827,
|
155 |
+
"grad_norm": 1.1261711120605469,
|
156 |
+
"learning_rate": 1.9800000000000004e-05,
|
157 |
+
"loss": 2.9798,
|
158 |
+
"step": 200
|
159 |
+
},
|
160 |
+
{
|
161 |
+
"epoch": 1.7241379310344827,
|
162 |
+
"eval_loss": 2.9736363887786865,
|
163 |
+
"eval_runtime": 39.6236,
|
164 |
+
"eval_samples_per_second": 33.919,
|
165 |
+
"eval_steps_per_second": 33.919,
|
166 |
+
"eval_wer": 1.0,
|
167 |
+
"step": 200
|
168 |
+
},
|
169 |
+
{
|
170 |
+
"epoch": 1.8103448275862069,
|
171 |
+
"grad_norm": 0.903380811214447,
|
172 |
+
"learning_rate": 2.08e-05,
|
173 |
+
"loss": 2.9718,
|
174 |
+
"step": 210
|
175 |
+
},
|
176 |
+
{
|
177 |
+
"epoch": 1.896551724137931,
|
178 |
+
"grad_norm": 0.4889620244503021,
|
179 |
+
"learning_rate": 2.18e-05,
|
180 |
+
"loss": 2.9766,
|
181 |
+
"step": 220
|
182 |
+
},
|
183 |
+
{
|
184 |
+
"epoch": 1.9827586206896552,
|
185 |
+
"grad_norm": 1.3861790895462036,
|
186 |
+
"learning_rate": 2.2800000000000002e-05,
|
187 |
+
"loss": 2.9658,
|
188 |
+
"step": 230
|
189 |
+
},
|
190 |
+
{
|
191 |
+
"epoch": 2.0689655172413794,
|
192 |
+
"grad_norm": 0.7976490259170532,
|
193 |
+
"learning_rate": 2.38e-05,
|
194 |
+
"loss": 2.9588,
|
195 |
+
"step": 240
|
196 |
+
},
|
197 |
+
{
|
198 |
+
"epoch": 2.1551724137931036,
|
199 |
+
"grad_norm": 0.698798418045044,
|
200 |
+
"learning_rate": 2.48e-05,
|
201 |
+
"loss": 2.9523,
|
202 |
+
"step": 250
|
203 |
+
},
|
204 |
+
{
|
205 |
+
"epoch": 2.2413793103448274,
|
206 |
+
"grad_norm": 1.0858148336410522,
|
207 |
+
"learning_rate": 2.58e-05,
|
208 |
+
"loss": 2.9496,
|
209 |
+
"step": 260
|
210 |
+
},
|
211 |
+
{
|
212 |
+
"epoch": 2.3275862068965516,
|
213 |
+
"grad_norm": 0.5658290386199951,
|
214 |
+
"learning_rate": 2.6800000000000004e-05,
|
215 |
+
"loss": 2.9421,
|
216 |
+
"step": 270
|
217 |
+
},
|
218 |
+
{
|
219 |
+
"epoch": 2.413793103448276,
|
220 |
+
"grad_norm": 0.5713534355163574,
|
221 |
+
"learning_rate": 2.7800000000000005e-05,
|
222 |
+
"loss": 2.9427,
|
223 |
+
"step": 280
|
224 |
+
},
|
225 |
+
{
|
226 |
+
"epoch": 2.5,
|
227 |
+
"grad_norm": 0.7386118769645691,
|
228 |
+
"learning_rate": 2.88e-05,
|
229 |
+
"loss": 2.9228,
|
230 |
+
"step": 290
|
231 |
+
},
|
232 |
+
{
|
233 |
+
"epoch": 2.586206896551724,
|
234 |
+
"grad_norm": 0.767816960811615,
|
235 |
+
"learning_rate": 2.98e-05,
|
236 |
+
"loss": 2.9144,
|
237 |
+
"step": 300
|
238 |
+
},
|
239 |
+
{
|
240 |
+
"epoch": 2.586206896551724,
|
241 |
+
"eval_loss": 2.9074809551239014,
|
242 |
+
"eval_runtime": 39.8997,
|
243 |
+
"eval_samples_per_second": 33.684,
|
244 |
+
"eval_steps_per_second": 33.684,
|
245 |
+
"eval_wer": 1.0,
|
246 |
+
"step": 300
|
247 |
+
},
|
248 |
+
{
|
249 |
+
"epoch": 2.6724137931034484,
|
250 |
+
"grad_norm": 0.8676608204841614,
|
251 |
+
"learning_rate": 3.08e-05,
|
252 |
+
"loss": 2.8965,
|
253 |
+
"step": 310
|
254 |
+
},
|
255 |
+
{
|
256 |
+
"epoch": 2.7586206896551726,
|
257 |
+
"grad_norm": 1.6954621076583862,
|
258 |
+
"learning_rate": 3.18e-05,
|
259 |
+
"loss": 2.8815,
|
260 |
+
"step": 320
|
261 |
+
},
|
262 |
+
{
|
263 |
+
"epoch": 2.844827586206897,
|
264 |
+
"grad_norm": 1.1631884574890137,
|
265 |
+
"learning_rate": 3.2800000000000004e-05,
|
266 |
+
"loss": 2.855,
|
267 |
+
"step": 330
|
268 |
+
},
|
269 |
+
{
|
270 |
+
"epoch": 2.9310344827586206,
|
271 |
+
"grad_norm": 1.625454306602478,
|
272 |
+
"learning_rate": 3.38e-05,
|
273 |
+
"loss": 2.781,
|
274 |
+
"step": 340
|
275 |
+
},
|
276 |
+
{
|
277 |
+
"epoch": 3.0172413793103448,
|
278 |
+
"grad_norm": 2.0763564109802246,
|
279 |
+
"learning_rate": 3.48e-05,
|
280 |
+
"loss": 2.7756,
|
281 |
+
"step": 350
|
282 |
+
},
|
283 |
+
{
|
284 |
+
"epoch": 3.103448275862069,
|
285 |
+
"grad_norm": 2.036031723022461,
|
286 |
+
"learning_rate": 3.58e-05,
|
287 |
+
"loss": 2.6458,
|
288 |
+
"step": 360
|
289 |
+
},
|
290 |
+
{
|
291 |
+
"epoch": 3.189655172413793,
|
292 |
+
"grad_norm": 1.366801142692566,
|
293 |
+
"learning_rate": 3.68e-05,
|
294 |
+
"loss": 2.5189,
|
295 |
+
"step": 370
|
296 |
+
},
|
297 |
+
{
|
298 |
+
"epoch": 3.2758620689655173,
|
299 |
+
"grad_norm": 2.034527540206909,
|
300 |
+
"learning_rate": 3.7800000000000004e-05,
|
301 |
+
"loss": 2.433,
|
302 |
+
"step": 380
|
303 |
+
},
|
304 |
+
{
|
305 |
+
"epoch": 3.3620689655172415,
|
306 |
+
"grad_norm": 3.8338165283203125,
|
307 |
+
"learning_rate": 3.88e-05,
|
308 |
+
"loss": 2.2885,
|
309 |
+
"step": 390
|
310 |
+
},
|
311 |
+
{
|
312 |
+
"epoch": 3.4482758620689653,
|
313 |
+
"grad_norm": 2.3443217277526855,
|
314 |
+
"learning_rate": 3.9800000000000005e-05,
|
315 |
+
"loss": 2.1714,
|
316 |
+
"step": 400
|
317 |
+
},
|
318 |
+
{
|
319 |
+
"epoch": 3.4482758620689653,
|
320 |
+
"eval_loss": 2.0944502353668213,
|
321 |
+
"eval_runtime": 39.7668,
|
322 |
+
"eval_samples_per_second": 33.797,
|
323 |
+
"eval_steps_per_second": 33.797,
|
324 |
+
"eval_wer": 1.0325047801147227,
|
325 |
+
"step": 400
|
326 |
+
},
|
327 |
+
{
|
328 |
+
"epoch": 3.5344827586206895,
|
329 |
+
"grad_norm": 4.349735260009766,
|
330 |
+
"learning_rate": 4.08e-05,
|
331 |
+
"loss": 2.0881,
|
332 |
+
"step": 410
|
333 |
+
},
|
334 |
+
{
|
335 |
+
"epoch": 3.6206896551724137,
|
336 |
+
"grad_norm": 2.450747489929199,
|
337 |
+
"learning_rate": 4.18e-05,
|
338 |
+
"loss": 1.9522,
|
339 |
+
"step": 420
|
340 |
+
},
|
341 |
+
{
|
342 |
+
"epoch": 3.706896551724138,
|
343 |
+
"grad_norm": 2.2519729137420654,
|
344 |
+
"learning_rate": 4.2800000000000004e-05,
|
345 |
+
"loss": 1.8395,
|
346 |
+
"step": 430
|
347 |
+
},
|
348 |
+
{
|
349 |
+
"epoch": 3.793103448275862,
|
350 |
+
"grad_norm": 2.693664789199829,
|
351 |
+
"learning_rate": 4.38e-05,
|
352 |
+
"loss": 1.7525,
|
353 |
+
"step": 440
|
354 |
+
},
|
355 |
+
{
|
356 |
+
"epoch": 3.8793103448275863,
|
357 |
+
"grad_norm": 1.9744929075241089,
|
358 |
+
"learning_rate": 4.4800000000000005e-05,
|
359 |
+
"loss": 1.6222,
|
360 |
+
"step": 450
|
361 |
+
},
|
362 |
+
{
|
363 |
+
"epoch": 3.9655172413793105,
|
364 |
+
"grad_norm": 3.802494764328003,
|
365 |
+
"learning_rate": 4.58e-05,
|
366 |
+
"loss": 1.5397,
|
367 |
+
"step": 460
|
368 |
+
},
|
369 |
+
{
|
370 |
+
"epoch": 4.051724137931035,
|
371 |
+
"grad_norm": 2.301044225692749,
|
372 |
+
"learning_rate": 4.6800000000000006e-05,
|
373 |
+
"loss": 1.4376,
|
374 |
+
"step": 470
|
375 |
+
},
|
376 |
+
{
|
377 |
+
"epoch": 4.137931034482759,
|
378 |
+
"grad_norm": 2.279372215270996,
|
379 |
+
"learning_rate": 4.78e-05,
|
380 |
+
"loss": 1.2829,
|
381 |
+
"step": 480
|
382 |
+
},
|
383 |
+
{
|
384 |
+
"epoch": 4.224137931034483,
|
385 |
+
"grad_norm": 3.314736843109131,
|
386 |
+
"learning_rate": 4.88e-05,
|
387 |
+
"loss": 1.1976,
|
388 |
+
"step": 490
|
389 |
+
},
|
390 |
+
{
|
391 |
+
"epoch": 4.310344827586207,
|
392 |
+
"grad_norm": 2.434694290161133,
|
393 |
+
"learning_rate": 4.9800000000000004e-05,
|
394 |
+
"loss": 1.1579,
|
395 |
+
"step": 500
|
396 |
+
},
|
397 |
+
{
|
398 |
+
"epoch": 4.310344827586207,
|
399 |
+
"eval_loss": 1.045101284980774,
|
400 |
+
"eval_runtime": 39.7455,
|
401 |
+
"eval_samples_per_second": 33.815,
|
402 |
+
"eval_steps_per_second": 33.815,
|
403 |
+
"eval_wer": 0.8299189656742239,
|
404 |
+
"step": 500
|
405 |
+
},
|
406 |
+
{
|
407 |
+
"epoch": 4.396551724137931,
|
408 |
+
"grad_norm": 1.8384031057357788,
|
409 |
+
"learning_rate": 5.08e-05,
|
410 |
+
"loss": 1.0684,
|
411 |
+
"step": 510
|
412 |
+
},
|
413 |
+
{
|
414 |
+
"epoch": 4.482758620689655,
|
415 |
+
"grad_norm": 3.599148988723755,
|
416 |
+
"learning_rate": 5.1800000000000005e-05,
|
417 |
+
"loss": 1.0319,
|
418 |
+
"step": 520
|
419 |
+
},
|
420 |
+
{
|
421 |
+
"epoch": 4.568965517241379,
|
422 |
+
"grad_norm": 2.066476583480835,
|
423 |
+
"learning_rate": 5.28e-05,
|
424 |
+
"loss": 0.9179,
|
425 |
+
"step": 530
|
426 |
+
},
|
427 |
+
{
|
428 |
+
"epoch": 4.655172413793103,
|
429 |
+
"grad_norm": 2.2173750400543213,
|
430 |
+
"learning_rate": 5.380000000000001e-05,
|
431 |
+
"loss": 0.8838,
|
432 |
+
"step": 540
|
433 |
+
},
|
434 |
+
{
|
435 |
+
"epoch": 4.741379310344827,
|
436 |
+
"grad_norm": 2.427091121673584,
|
437 |
+
"learning_rate": 5.4800000000000004e-05,
|
438 |
+
"loss": 0.8991,
|
439 |
+
"step": 550
|
440 |
+
},
|
441 |
+
{
|
442 |
+
"epoch": 4.827586206896552,
|
443 |
+
"grad_norm": 2.7432241439819336,
|
444 |
+
"learning_rate": 5.580000000000001e-05,
|
445 |
+
"loss": 0.8,
|
446 |
+
"step": 560
|
447 |
+
},
|
448 |
+
{
|
449 |
+
"epoch": 4.913793103448276,
|
450 |
+
"grad_norm": 3.254221200942993,
|
451 |
+
"learning_rate": 5.68e-05,
|
452 |
+
"loss": 0.7803,
|
453 |
+
"step": 570
|
454 |
+
},
|
455 |
+
{
|
456 |
+
"epoch": 5.0,
|
457 |
+
"grad_norm": 4.457448482513428,
|
458 |
+
"learning_rate": 5.7799999999999995e-05,
|
459 |
+
"loss": 0.8205,
|
460 |
+
"step": 580
|
461 |
+
},
|
462 |
+
{
|
463 |
+
"epoch": 5.086206896551724,
|
464 |
+
"grad_norm": 3.1023166179656982,
|
465 |
+
"learning_rate": 5.88e-05,
|
466 |
+
"loss": 0.6703,
|
467 |
+
"step": 590
|
468 |
+
},
|
469 |
+
{
|
470 |
+
"epoch": 5.172413793103448,
|
471 |
+
"grad_norm": 2.5916504859924316,
|
472 |
+
"learning_rate": 5.9800000000000003e-05,
|
473 |
+
"loss": 0.6087,
|
474 |
+
"step": 600
|
475 |
+
},
|
476 |
+
{
|
477 |
+
"epoch": 5.172413793103448,
|
478 |
+
"eval_loss": 0.6753795146942139,
|
479 |
+
"eval_runtime": 39.7485,
|
480 |
+
"eval_samples_per_second": 33.813,
|
481 |
+
"eval_steps_per_second": 33.813,
|
482 |
+
"eval_wer": 0.6440863152144223,
|
483 |
+
"step": 600
|
484 |
+
},
|
485 |
+
{
|
486 |
+
"epoch": 5.258620689655173,
|
487 |
+
"grad_norm": 2.1707613468170166,
|
488 |
+
"learning_rate": 6.08e-05,
|
489 |
+
"loss": 0.6569,
|
490 |
+
"step": 610
|
491 |
+
},
|
492 |
+
{
|
493 |
+
"epoch": 5.344827586206897,
|
494 |
+
"grad_norm": 2.4291555881500244,
|
495 |
+
"learning_rate": 6.18e-05,
|
496 |
+
"loss": 0.5627,
|
497 |
+
"step": 620
|
498 |
+
},
|
499 |
+
{
|
500 |
+
"epoch": 5.431034482758621,
|
501 |
+
"grad_norm": 2.249617338180542,
|
502 |
+
"learning_rate": 6.280000000000001e-05,
|
503 |
+
"loss": 0.5381,
|
504 |
+
"step": 630
|
505 |
+
},
|
506 |
+
{
|
507 |
+
"epoch": 5.517241379310345,
|
508 |
+
"grad_norm": 1.6661946773529053,
|
509 |
+
"learning_rate": 6.38e-05,
|
510 |
+
"loss": 0.6338,
|
511 |
+
"step": 640
|
512 |
+
},
|
513 |
+
{
|
514 |
+
"epoch": 5.603448275862069,
|
515 |
+
"grad_norm": 2.60294771194458,
|
516 |
+
"learning_rate": 6.48e-05,
|
517 |
+
"loss": 0.5181,
|
518 |
+
"step": 650
|
519 |
+
},
|
520 |
+
{
|
521 |
+
"epoch": 5.689655172413794,
|
522 |
+
"grad_norm": 3.3003089427948,
|
523 |
+
"learning_rate": 6.58e-05,
|
524 |
+
"loss": 0.5189,
|
525 |
+
"step": 660
|
526 |
+
},
|
527 |
+
{
|
528 |
+
"epoch": 5.775862068965517,
|
529 |
+
"grad_norm": 1.880764126777649,
|
530 |
+
"learning_rate": 6.680000000000001e-05,
|
531 |
+
"loss": 0.564,
|
532 |
+
"step": 670
|
533 |
+
},
|
534 |
+
{
|
535 |
+
"epoch": 5.862068965517241,
|
536 |
+
"grad_norm": 2.0575127601623535,
|
537 |
+
"learning_rate": 6.780000000000001e-05,
|
538 |
+
"loss": 0.4729,
|
539 |
+
"step": 680
|
540 |
+
},
|
541 |
+
{
|
542 |
+
"epoch": 5.948275862068965,
|
543 |
+
"grad_norm": 2.5159761905670166,
|
544 |
+
"learning_rate": 6.879999999999999e-05,
|
545 |
+
"loss": 0.4899,
|
546 |
+
"step": 690
|
547 |
+
},
|
548 |
+
{
|
549 |
+
"epoch": 6.0344827586206895,
|
550 |
+
"grad_norm": 1.4463504552841187,
|
551 |
+
"learning_rate": 6.98e-05,
|
552 |
+
"loss": 0.481,
|
553 |
+
"step": 700
|
554 |
+
},
|
555 |
+
{
|
556 |
+
"epoch": 6.0344827586206895,
|
557 |
+
"eval_loss": 0.5275412201881409,
|
558 |
+
"eval_runtime": 39.9601,
|
559 |
+
"eval_samples_per_second": 33.634,
|
560 |
+
"eval_steps_per_second": 33.634,
|
561 |
+
"eval_wer": 0.5760721114449604,
|
562 |
+
"step": 700
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 6.120689655172414,
|
566 |
+
"grad_norm": 1.788765549659729,
|
567 |
+
"learning_rate": 7.08e-05,
|
568 |
+
"loss": 0.3865,
|
569 |
+
"step": 710
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 6.206896551724138,
|
573 |
+
"grad_norm": 1.862762212753296,
|
574 |
+
"learning_rate": 7.18e-05,
|
575 |
+
"loss": 0.3726,
|
576 |
+
"step": 720
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 6.293103448275862,
|
580 |
+
"grad_norm": 1.6512093544006348,
|
581 |
+
"learning_rate": 7.280000000000001e-05,
|
582 |
+
"loss": 0.4116,
|
583 |
+
"step": 730
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 6.379310344827586,
|
587 |
+
"grad_norm": 2.098067045211792,
|
588 |
+
"learning_rate": 7.38e-05,
|
589 |
+
"loss": 0.3779,
|
590 |
+
"step": 740
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 6.4655172413793105,
|
594 |
+
"grad_norm": 3.3030078411102295,
|
595 |
+
"learning_rate": 7.48e-05,
|
596 |
+
"loss": 0.3728,
|
597 |
+
"step": 750
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 6.551724137931035,
|
601 |
+
"grad_norm": 2.1799120903015137,
|
602 |
+
"learning_rate": 7.58e-05,
|
603 |
+
"loss": 0.4047,
|
604 |
+
"step": 760
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 6.637931034482759,
|
608 |
+
"grad_norm": 1.862434983253479,
|
609 |
+
"learning_rate": 7.680000000000001e-05,
|
610 |
+
"loss": 0.313,
|
611 |
+
"step": 770
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 6.724137931034483,
|
615 |
+
"grad_norm": 6.29113245010376,
|
616 |
+
"learning_rate": 7.780000000000001e-05,
|
617 |
+
"loss": 0.4052,
|
618 |
+
"step": 780
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 6.810344827586206,
|
622 |
+
"grad_norm": 1.4220325946807861,
|
623 |
+
"learning_rate": 7.88e-05,
|
624 |
+
"loss": 0.3218,
|
625 |
+
"step": 790
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 6.896551724137931,
|
629 |
+
"grad_norm": 2.586819648742676,
|
630 |
+
"learning_rate": 7.98e-05,
|
631 |
+
"loss": 0.3072,
|
632 |
+
"step": 800
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 6.896551724137931,
|
636 |
+
"eval_loss": 0.4836220443248749,
|
637 |
+
"eval_runtime": 39.8762,
|
638 |
+
"eval_samples_per_second": 33.704,
|
639 |
+
"eval_steps_per_second": 33.704,
|
640 |
+
"eval_wer": 0.5264499681325685,
|
641 |
+
"step": 800
|
642 |
+
},
|
643 |
+
{
|
644 |
+
"epoch": 6.982758620689655,
|
645 |
+
"grad_norm": 1.6589460372924805,
|
646 |
+
"learning_rate": 8.080000000000001e-05,
|
647 |
+
"loss": 0.3862,
|
648 |
+
"step": 810
|
649 |
+
},
|
650 |
+
{
|
651 |
+
"epoch": 7.068965517241379,
|
652 |
+
"grad_norm": 1.7299175262451172,
|
653 |
+
"learning_rate": 8.18e-05,
|
654 |
+
"loss": 0.2938,
|
655 |
+
"step": 820
|
656 |
+
},
|
657 |
+
{
|
658 |
+
"epoch": 7.155172413793103,
|
659 |
+
"grad_norm": 2.0545098781585693,
|
660 |
+
"learning_rate": 8.28e-05,
|
661 |
+
"loss": 0.249,
|
662 |
+
"step": 830
|
663 |
+
},
|
664 |
+
{
|
665 |
+
"epoch": 7.241379310344827,
|
666 |
+
"grad_norm": 24.935670852661133,
|
667 |
+
"learning_rate": 8.38e-05,
|
668 |
+
"loss": 0.3202,
|
669 |
+
"step": 840
|
670 |
+
},
|
671 |
+
{
|
672 |
+
"epoch": 7.327586206896552,
|
673 |
+
"grad_norm": 2.497840642929077,
|
674 |
+
"learning_rate": 8.48e-05,
|
675 |
+
"loss": 0.2803,
|
676 |
+
"step": 850
|
677 |
+
},
|
678 |
+
{
|
679 |
+
"epoch": 7.413793103448276,
|
680 |
+
"grad_norm": 2.698636531829834,
|
681 |
+
"learning_rate": 8.58e-05,
|
682 |
+
"loss": 0.2473,
|
683 |
+
"step": 860
|
684 |
+
},
|
685 |
+
{
|
686 |
+
"epoch": 7.5,
|
687 |
+
"grad_norm": 1.4561227560043335,
|
688 |
+
"learning_rate": 8.680000000000001e-05,
|
689 |
+
"loss": 0.3223,
|
690 |
+
"step": 870
|
691 |
+
},
|
692 |
+
{
|
693 |
+
"epoch": 7.586206896551724,
|
694 |
+
"grad_norm": 1.7760556936264038,
|
695 |
+
"learning_rate": 8.78e-05,
|
696 |
+
"loss": 0.2481,
|
697 |
+
"step": 880
|
698 |
+
},
|
699 |
+
{
|
700 |
+
"epoch": 7.672413793103448,
|
701 |
+
"grad_norm": 2.308103084564209,
|
702 |
+
"learning_rate": 8.88e-05,
|
703 |
+
"loss": 0.2545,
|
704 |
+
"step": 890
|
705 |
+
},
|
706 |
+
{
|
707 |
+
"epoch": 7.758620689655173,
|
708 |
+
"grad_norm": 1.4128385782241821,
|
709 |
+
"learning_rate": 8.98e-05,
|
710 |
+
"loss": 0.332,
|
711 |
+
"step": 900
|
712 |
+
},
|
713 |
+
{
|
714 |
+
"epoch": 7.758620689655173,
|
715 |
+
"eval_loss": 0.44030094146728516,
|
716 |
+
"eval_runtime": 39.9401,
|
717 |
+
"eval_samples_per_second": 33.65,
|
718 |
+
"eval_steps_per_second": 33.65,
|
719 |
+
"eval_wer": 0.5233542747883092,
|
720 |
+
"step": 900
|
721 |
+
},
|
722 |
+
{
|
723 |
+
"epoch": 7.844827586206897,
|
724 |
+
"grad_norm": 1.7903906106948853,
|
725 |
+
"learning_rate": 9.080000000000001e-05,
|
726 |
+
"loss": 0.2411,
|
727 |
+
"step": 910
|
728 |
+
},
|
729 |
+
{
|
730 |
+
"epoch": 7.931034482758621,
|
731 |
+
"grad_norm": 2.0804216861724854,
|
732 |
+
"learning_rate": 9.180000000000001e-05,
|
733 |
+
"loss": 0.2707,
|
734 |
+
"step": 920
|
735 |
+
},
|
736 |
+
{
|
737 |
+
"epoch": 8.017241379310345,
|
738 |
+
"grad_norm": 1.4420605897903442,
|
739 |
+
"learning_rate": 9.28e-05,
|
740 |
+
"loss": 0.3186,
|
741 |
+
"step": 930
|
742 |
+
},
|
743 |
+
{
|
744 |
+
"epoch": 8.10344827586207,
|
745 |
+
"grad_norm": 2.2910854816436768,
|
746 |
+
"learning_rate": 9.38e-05,
|
747 |
+
"loss": 0.1937,
|
748 |
+
"step": 940
|
749 |
+
},
|
750 |
+
{
|
751 |
+
"epoch": 8.189655172413794,
|
752 |
+
"grad_norm": 3.5892796516418457,
|
753 |
+
"learning_rate": 9.48e-05,
|
754 |
+
"loss": 0.2321,
|
755 |
+
"step": 950
|
756 |
+
},
|
757 |
+
{
|
758 |
+
"epoch": 8.275862068965518,
|
759 |
+
"grad_norm": 1.6509956121444702,
|
760 |
+
"learning_rate": 9.58e-05,
|
761 |
+
"loss": 0.2868,
|
762 |
+
"step": 960
|
763 |
+
},
|
764 |
+
{
|
765 |
+
"epoch": 8.362068965517242,
|
766 |
+
"grad_norm": 1.6983604431152344,
|
767 |
+
"learning_rate": 9.680000000000001e-05,
|
768 |
+
"loss": 0.2004,
|
769 |
+
"step": 970
|
770 |
+
},
|
771 |
+
{
|
772 |
+
"epoch": 8.448275862068966,
|
773 |
+
"grad_norm": 2.061176061630249,
|
774 |
+
"learning_rate": 9.78e-05,
|
775 |
+
"loss": 0.2025,
|
776 |
+
"step": 980
|
777 |
+
},
|
778 |
+
{
|
779 |
+
"epoch": 8.53448275862069,
|
780 |
+
"grad_norm": 1.7732270956039429,
|
781 |
+
"learning_rate": 9.88e-05,
|
782 |
+
"loss": 0.2598,
|
783 |
+
"step": 990
|
784 |
+
},
|
785 |
+
{
|
786 |
+
"epoch": 8.620689655172415,
|
787 |
+
"grad_norm": 1.8335466384887695,
|
788 |
+
"learning_rate": 9.98e-05,
|
789 |
+
"loss": 0.1876,
|
790 |
+
"step": 1000
|
791 |
+
},
|
792 |
+
{
|
793 |
+
"epoch": 8.620689655172415,
|
794 |
+
"eval_loss": 0.4757933020591736,
|
795 |
+
"eval_runtime": 39.8291,
|
796 |
+
"eval_samples_per_second": 33.744,
|
797 |
+
"eval_steps_per_second": 33.744,
|
798 |
+
"eval_wer": 0.5221706273331512,
|
799 |
+
"step": 1000
|
800 |
+
},
|
801 |
+
{
|
802 |
+
"epoch": 8.706896551724139,
|
803 |
+
"grad_norm": 2.52902889251709,
|
804 |
+
"learning_rate": 9.939393939393939e-05,
|
805 |
+
"loss": 0.2456,
|
806 |
+
"step": 1010
|
807 |
+
},
|
808 |
+
{
|
809 |
+
"epoch": 8.793103448275861,
|
810 |
+
"grad_norm": 1.7294162511825562,
|
811 |
+
"learning_rate": 9.863636363636364e-05,
|
812 |
+
"loss": 0.2499,
|
813 |
+
"step": 1020
|
814 |
+
},
|
815 |
+
{
|
816 |
+
"epoch": 8.879310344827585,
|
817 |
+
"grad_norm": 21.9121150970459,
|
818 |
+
"learning_rate": 9.787878787878789e-05,
|
819 |
+
"loss": 0.1854,
|
820 |
+
"step": 1030
|
821 |
+
},
|
822 |
+
{
|
823 |
+
"epoch": 8.96551724137931,
|
824 |
+
"grad_norm": 3.9164559841156006,
|
825 |
+
"learning_rate": 9.712121212121212e-05,
|
826 |
+
"loss": 0.2576,
|
827 |
+
"step": 1040
|
828 |
+
},
|
829 |
+
{
|
830 |
+
"epoch": 9.051724137931034,
|
831 |
+
"grad_norm": 1.239221215248108,
|
832 |
+
"learning_rate": 9.636363636363637e-05,
|
833 |
+
"loss": 0.2118,
|
834 |
+
"step": 1050
|
835 |
+
},
|
836 |
+
{
|
837 |
+
"epoch": 9.137931034482758,
|
838 |
+
"grad_norm": 3.1416544914245605,
|
839 |
+
"learning_rate": 9.560606060606061e-05,
|
840 |
+
"loss": 0.1577,
|
841 |
+
"step": 1060
|
842 |
+
},
|
843 |
+
{
|
844 |
+
"epoch": 9.224137931034482,
|
845 |
+
"grad_norm": 2.4253621101379395,
|
846 |
+
"learning_rate": 9.484848484848486e-05,
|
847 |
+
"loss": 0.2092,
|
848 |
+
"step": 1070
|
849 |
+
},
|
850 |
+
{
|
851 |
+
"epoch": 9.310344827586206,
|
852 |
+
"grad_norm": 1.194345474243164,
|
853 |
+
"learning_rate": 9.40909090909091e-05,
|
854 |
+
"loss": 0.1876,
|
855 |
+
"step": 1080
|
856 |
+
},
|
857 |
+
{
|
858 |
+
"epoch": 9.39655172413793,
|
859 |
+
"grad_norm": 2.411029100418091,
|
860 |
+
"learning_rate": 9.333333333333334e-05,
|
861 |
+
"loss": 0.1546,
|
862 |
+
"step": 1090
|
863 |
+
},
|
864 |
+
{
|
865 |
+
"epoch": 9.482758620689655,
|
866 |
+
"grad_norm": 3.246082067489624,
|
867 |
+
"learning_rate": 9.257575757575758e-05,
|
868 |
+
"loss": 0.2232,
|
869 |
+
"step": 1100
|
870 |
+
},
|
871 |
+
{
|
872 |
+
"epoch": 9.482758620689655,
|
873 |
+
"eval_loss": 0.45077577233314514,
|
874 |
+
"eval_runtime": 39.9221,
|
875 |
+
"eval_samples_per_second": 33.666,
|
876 |
+
"eval_steps_per_second": 33.666,
|
877 |
+
"eval_wer": 0.48921059819721385,
|
878 |
+
"step": 1100
|
879 |
+
},
|
880 |
+
{
|
881 |
+
"epoch": 9.568965517241379,
|
882 |
+
"grad_norm": 1.3427454233169556,
|
883 |
+
"learning_rate": 9.181818181818183e-05,
|
884 |
+
"loss": 0.1777,
|
885 |
+
"step": 1110
|
886 |
+
},
|
887 |
+
{
|
888 |
+
"epoch": 9.655172413793103,
|
889 |
+
"grad_norm": 1.5090447664260864,
|
890 |
+
"learning_rate": 9.106060606060606e-05,
|
891 |
+
"loss": 0.1646,
|
892 |
+
"step": 1120
|
893 |
+
},
|
894 |
+
{
|
895 |
+
"epoch": 9.741379310344827,
|
896 |
+
"grad_norm": 1.3060975074768066,
|
897 |
+
"learning_rate": 9.030303030303031e-05,
|
898 |
+
"loss": 0.225,
|
899 |
+
"step": 1130
|
900 |
+
},
|
901 |
+
{
|
902 |
+
"epoch": 9.827586206896552,
|
903 |
+
"grad_norm": 1.3011540174484253,
|
904 |
+
"learning_rate": 8.954545454545455e-05,
|
905 |
+
"loss": 0.1552,
|
906 |
+
"step": 1140
|
907 |
+
},
|
908 |
+
{
|
909 |
+
"epoch": 9.913793103448276,
|
910 |
+
"grad_norm": 1.9938538074493408,
|
911 |
+
"learning_rate": 8.87878787878788e-05,
|
912 |
+
"loss": 0.1715,
|
913 |
+
"step": 1150
|
914 |
+
},
|
915 |
+
{
|
916 |
+
"epoch": 10.0,
|
917 |
+
"grad_norm": 3.334385395050049,
|
918 |
+
"learning_rate": 8.803030303030304e-05,
|
919 |
+
"loss": 0.2092,
|
920 |
+
"step": 1160
|
921 |
+
},
|
922 |
+
{
|
923 |
+
"epoch": 10.086206896551724,
|
924 |
+
"grad_norm": 1.011092185974121,
|
925 |
+
"learning_rate": 8.727272727272727e-05,
|
926 |
+
"loss": 0.14,
|
927 |
+
"step": 1170
|
928 |
+
},
|
929 |
+
{
|
930 |
+
"epoch": 10.172413793103448,
|
931 |
+
"grad_norm": 2.517902135848999,
|
932 |
+
"learning_rate": 8.651515151515152e-05,
|
933 |
+
"loss": 0.1512,
|
934 |
+
"step": 1180
|
935 |
+
},
|
936 |
+
{
|
937 |
+
"epoch": 10.258620689655173,
|
938 |
+
"grad_norm": 1.2418378591537476,
|
939 |
+
"learning_rate": 8.575757575757576e-05,
|
940 |
+
"loss": 0.1846,
|
941 |
+
"step": 1190
|
942 |
+
},
|
943 |
+
{
|
944 |
+
"epoch": 10.344827586206897,
|
945 |
+
"grad_norm": 1.5885329246520996,
|
946 |
+
"learning_rate": 8.5e-05,
|
947 |
+
"loss": 0.1332,
|
948 |
+
"step": 1200
|
949 |
+
},
|
950 |
+
{
|
951 |
+
"epoch": 10.344827586206897,
|
952 |
+
"eval_loss": 0.4394075274467468,
|
953 |
+
"eval_runtime": 39.9367,
|
954 |
+
"eval_samples_per_second": 33.653,
|
955 |
+
"eval_steps_per_second": 33.653,
|
956 |
+
"eval_wer": 0.4740052808886461,
|
957 |
+
"step": 1200
|
958 |
+
},
|
959 |
+
{
|
960 |
+
"epoch": 10.431034482758621,
|
961 |
+
"grad_norm": 1.2539469003677368,
|
962 |
+
"learning_rate": 8.424242424242424e-05,
|
963 |
+
"loss": 0.1485,
|
964 |
+
"step": 1210
|
965 |
+
},
|
966 |
+
{
|
967 |
+
"epoch": 10.517241379310345,
|
968 |
+
"grad_norm": 1.357601284980774,
|
969 |
+
"learning_rate": 8.348484848484849e-05,
|
970 |
+
"loss": 0.1988,
|
971 |
+
"step": 1220
|
972 |
+
},
|
973 |
+
{
|
974 |
+
"epoch": 10.60344827586207,
|
975 |
+
"grad_norm": 2.0564587116241455,
|
976 |
+
"learning_rate": 8.272727272727273e-05,
|
977 |
+
"loss": 0.137,
|
978 |
+
"step": 1230
|
979 |
+
},
|
980 |
+
{
|
981 |
+
"epoch": 10.689655172413794,
|
982 |
+
"grad_norm": 2.48364520072937,
|
983 |
+
"learning_rate": 8.196969696969698e-05,
|
984 |
+
"loss": 0.1245,
|
985 |
+
"step": 1240
|
986 |
+
},
|
987 |
+
{
|
988 |
+
"epoch": 10.775862068965518,
|
989 |
+
"grad_norm": 1.015891671180725,
|
990 |
+
"learning_rate": 8.121212121212121e-05,
|
991 |
+
"loss": 0.1602,
|
992 |
+
"step": 1250
|
993 |
+
},
|
994 |
+
{
|
995 |
+
"epoch": 10.862068965517242,
|
996 |
+
"grad_norm": 1.1023950576782227,
|
997 |
+
"learning_rate": 8.045454545454546e-05,
|
998 |
+
"loss": 0.1215,
|
999 |
+
"step": 1260
|
1000 |
+
},
|
1001 |
+
{
|
1002 |
+
"epoch": 10.948275862068966,
|
1003 |
+
"grad_norm": 2.703427791595459,
|
1004 |
+
"learning_rate": 7.96969696969697e-05,
|
1005 |
+
"loss": 0.1621,
|
1006 |
+
"step": 1270
|
1007 |
+
},
|
1008 |
+
{
|
1009 |
+
"epoch": 11.03448275862069,
|
1010 |
+
"grad_norm": 1.1821691989898682,
|
1011 |
+
"learning_rate": 7.893939393939395e-05,
|
1012 |
+
"loss": 0.1651,
|
1013 |
+
"step": 1280
|
1014 |
+
},
|
1015 |
+
{
|
1016 |
+
"epoch": 11.120689655172415,
|
1017 |
+
"grad_norm": 0.930283784866333,
|
1018 |
+
"learning_rate": 7.818181818181818e-05,
|
1019 |
+
"loss": 0.1066,
|
1020 |
+
"step": 1290
|
1021 |
+
},
|
1022 |
+
{
|
1023 |
+
"epoch": 11.206896551724139,
|
1024 |
+
"grad_norm": 1.6548758745193481,
|
1025 |
+
"learning_rate": 7.742424242424243e-05,
|
1026 |
+
"loss": 0.1085,
|
1027 |
+
"step": 1300
|
1028 |
+
},
|
1029 |
+
{
|
1030 |
+
"epoch": 11.206896551724139,
|
1031 |
+
"eval_loss": 0.4466467499732971,
|
1032 |
+
"eval_runtime": 39.8633,
|
1033 |
+
"eval_samples_per_second": 33.715,
|
1034 |
+
"eval_steps_per_second": 33.715,
|
1035 |
+
"eval_wer": 0.46207775653282346,
|
1036 |
+
"step": 1300
|
1037 |
+
},
|
1038 |
+
{
|
1039 |
+
"epoch": 11.293103448275861,
|
1040 |
+
"grad_norm": 1.1760716438293457,
|
1041 |
+
"learning_rate": 7.666666666666667e-05,
|
1042 |
+
"loss": 0.1418,
|
1043 |
+
"step": 1310
|
1044 |
+
},
|
1045 |
+
{
|
1046 |
+
"epoch": 11.379310344827585,
|
1047 |
+
"grad_norm": 2.1062755584716797,
|
1048 |
+
"learning_rate": 7.59090909090909e-05,
|
1049 |
+
"loss": 0.1133,
|
1050 |
+
"step": 1320
|
1051 |
+
},
|
1052 |
+
{
|
1053 |
+
"epoch": 11.46551724137931,
|
1054 |
+
"grad_norm": 2.67399001121521,
|
1055 |
+
"learning_rate": 7.515151515151515e-05,
|
1056 |
+
"loss": 0.1318,
|
1057 |
+
"step": 1330
|
1058 |
+
},
|
1059 |
+
{
|
1060 |
+
"epoch": 11.551724137931034,
|
1061 |
+
"grad_norm": 1.0049142837524414,
|
1062 |
+
"learning_rate": 7.439393939393939e-05,
|
1063 |
+
"loss": 0.1474,
|
1064 |
+
"step": 1340
|
1065 |
+
},
|
1066 |
+
{
|
1067 |
+
"epoch": 11.637931034482758,
|
1068 |
+
"grad_norm": 1.586559772491455,
|
1069 |
+
"learning_rate": 7.363636363636364e-05,
|
1070 |
+
"loss": 0.0908,
|
1071 |
+
"step": 1350
|
1072 |
+
},
|
1073 |
+
{
|
1074 |
+
"epoch": 11.724137931034482,
|
1075 |
+
"grad_norm": 3.784040927886963,
|
1076 |
+
"learning_rate": 7.287878787878788e-05,
|
1077 |
+
"loss": 0.1521,
|
1078 |
+
"step": 1360
|
1079 |
+
},
|
1080 |
+
{
|
1081 |
+
"epoch": 11.810344827586206,
|
1082 |
+
"grad_norm": 1.125501275062561,
|
1083 |
+
"learning_rate": 7.212121212121213e-05,
|
1084 |
+
"loss": 0.1163,
|
1085 |
+
"step": 1370
|
1086 |
+
},
|
1087 |
+
{
|
1088 |
+
"epoch": 11.89655172413793,
|
1089 |
+
"grad_norm": 2.1989808082580566,
|
1090 |
+
"learning_rate": 7.136363636363636e-05,
|
1091 |
+
"loss": 0.1109,
|
1092 |
+
"step": 1380
|
1093 |
+
},
|
1094 |
+
{
|
1095 |
+
"epoch": 11.982758620689655,
|
1096 |
+
"grad_norm": 1.1287301778793335,
|
1097 |
+
"learning_rate": 7.060606060606061e-05,
|
1098 |
+
"loss": 0.152,
|
1099 |
+
"step": 1390
|
1100 |
+
},
|
1101 |
+
{
|
1102 |
+
"epoch": 12.068965517241379,
|
1103 |
+
"grad_norm": 1.538678765296936,
|
1104 |
+
"learning_rate": 6.984848484848485e-05,
|
1105 |
+
"loss": 0.098,
|
1106 |
+
"step": 1400
|
1107 |
+
},
|
1108 |
+
{
|
1109 |
+
"epoch": 12.068965517241379,
|
1110 |
+
"eval_loss": 0.42302384972572327,
|
1111 |
+
"eval_runtime": 40.1773,
|
1112 |
+
"eval_samples_per_second": 33.452,
|
1113 |
+
"eval_steps_per_second": 33.452,
|
1114 |
+
"eval_wer": 0.44933078393881454,
|
1115 |
+
"step": 1400
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 12.155172413793103,
|
1119 |
+
"grad_norm": 1.400772213935852,
|
1120 |
+
"learning_rate": 6.90909090909091e-05,
|
1121 |
+
"loss": 0.092,
|
1122 |
+
"step": 1410
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 12.241379310344827,
|
1126 |
+
"grad_norm": 3.6780846118927,
|
1127 |
+
"learning_rate": 6.833333333333333e-05,
|
1128 |
+
"loss": 0.1649,
|
1129 |
+
"step": 1420
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 12.327586206896552,
|
1133 |
+
"grad_norm": 1.5424057245254517,
|
1134 |
+
"learning_rate": 6.757575757575758e-05,
|
1135 |
+
"loss": 0.091,
|
1136 |
+
"step": 1430
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 12.413793103448276,
|
1140 |
+
"grad_norm": 1.4868180751800537,
|
1141 |
+
"learning_rate": 6.681818181818183e-05,
|
1142 |
+
"loss": 0.0869,
|
1143 |
+
"step": 1440
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 12.5,
|
1147 |
+
"grad_norm": 1.1947145462036133,
|
1148 |
+
"learning_rate": 6.606060606060607e-05,
|
1149 |
+
"loss": 0.1499,
|
1150 |
+
"step": 1450
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 12.586206896551724,
|
1154 |
+
"grad_norm": 1.0430784225463867,
|
1155 |
+
"learning_rate": 6.530303030303032e-05,
|
1156 |
+
"loss": 0.0954,
|
1157 |
+
"step": 1460
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 12.672413793103448,
|
1161 |
+
"grad_norm": 2.4261584281921387,
|
1162 |
+
"learning_rate": 6.454545454545455e-05,
|
1163 |
+
"loss": 0.1032,
|
1164 |
+
"step": 1470
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 12.758620689655173,
|
1168 |
+
"grad_norm": 1.033467411994934,
|
1169 |
+
"learning_rate": 6.37878787878788e-05,
|
1170 |
+
"loss": 0.1158,
|
1171 |
+
"step": 1480
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 12.844827586206897,
|
1175 |
+
"grad_norm": 1.1535651683807373,
|
1176 |
+
"learning_rate": 6.303030303030302e-05,
|
1177 |
+
"loss": 0.0864,
|
1178 |
+
"step": 1490
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 12.931034482758621,
|
1182 |
+
"grad_norm": 1.28826105594635,
|
1183 |
+
"learning_rate": 6.227272727272727e-05,
|
1184 |
+
"loss": 0.1219,
|
1185 |
+
"step": 1500
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 12.931034482758621,
|
1189 |
+
"eval_loss": 0.418023020029068,
|
1190 |
+
"eval_runtime": 40.2192,
|
1191 |
+
"eval_samples_per_second": 33.417,
|
1192 |
+
"eval_steps_per_second": 33.417,
|
1193 |
+
"eval_wer": 0.44596194118182647,
|
1194 |
+
"step": 1500
|
1195 |
+
},
|
1196 |
+
{
|
1197 |
+
"epoch": 13.017241379310345,
|
1198 |
+
"grad_norm": 1.055411458015442,
|
1199 |
+
"learning_rate": 6.151515151515151e-05,
|
1200 |
+
"loss": 0.1289,
|
1201 |
+
"step": 1510
|
1202 |
+
},
|
1203 |
+
{
|
1204 |
+
"epoch": 13.10344827586207,
|
1205 |
+
"grad_norm": 1.1269094944000244,
|
1206 |
+
"learning_rate": 6.075757575757576e-05,
|
1207 |
+
"loss": 0.0776,
|
1208 |
+
"step": 1520
|
1209 |
+
},
|
1210 |
+
{
|
1211 |
+
"epoch": 13.189655172413794,
|
1212 |
+
"grad_norm": 1.7149118185043335,
|
1213 |
+
"learning_rate": 6e-05,
|
1214 |
+
"loss": 0.0871,
|
1215 |
+
"step": 1530
|
1216 |
+
},
|
1217 |
+
{
|
1218 |
+
"epoch": 13.275862068965518,
|
1219 |
+
"grad_norm": 1.7456856966018677,
|
1220 |
+
"learning_rate": 5.9242424242424244e-05,
|
1221 |
+
"loss": 0.1087,
|
1222 |
+
"step": 1540
|
1223 |
+
},
|
1224 |
+
{
|
1225 |
+
"epoch": 13.362068965517242,
|
1226 |
+
"grad_norm": 1.3434715270996094,
|
1227 |
+
"learning_rate": 5.848484848484849e-05,
|
1228 |
+
"loss": 0.0821,
|
1229 |
+
"step": 1550
|
1230 |
+
},
|
1231 |
+
{
|
1232 |
+
"epoch": 13.448275862068966,
|
1233 |
+
"grad_norm": 2.103512763977051,
|
1234 |
+
"learning_rate": 5.772727272727273e-05,
|
1235 |
+
"loss": 0.0878,
|
1236 |
+
"step": 1560
|
1237 |
+
},
|
1238 |
+
{
|
1239 |
+
"epoch": 13.53448275862069,
|
1240 |
+
"grad_norm": 1.240224838256836,
|
1241 |
+
"learning_rate": 5.696969696969697e-05,
|
1242 |
+
"loss": 0.1044,
|
1243 |
+
"step": 1570
|
1244 |
+
},
|
1245 |
+
{
|
1246 |
+
"epoch": 13.620689655172415,
|
1247 |
+
"grad_norm": 0.7336703538894653,
|
1248 |
+
"learning_rate": 5.6212121212121215e-05,
|
1249 |
+
"loss": 0.0753,
|
1250 |
+
"step": 1580
|
1251 |
+
},
|
1252 |
+
{
|
1253 |
+
"epoch": 13.706896551724139,
|
1254 |
+
"grad_norm": 2.293342351913452,
|
1255 |
+
"learning_rate": 5.545454545454546e-05,
|
1256 |
+
"loss": 0.1059,
|
1257 |
+
"step": 1590
|
1258 |
+
},
|
1259 |
+
{
|
1260 |
+
"epoch": 13.793103448275861,
|
1261 |
+
"grad_norm": 1.1853971481323242,
|
1262 |
+
"learning_rate": 5.46969696969697e-05,
|
1263 |
+
"loss": 0.1021,
|
1264 |
+
"step": 1600
|
1265 |
+
},
|
1266 |
+
{
|
1267 |
+
"epoch": 13.793103448275861,
|
1268 |
+
"eval_loss": 0.41785839200019836,
|
1269 |
+
"eval_runtime": 40.2906,
|
1270 |
+
"eval_samples_per_second": 33.358,
|
1271 |
+
"eval_steps_per_second": 33.358,
|
1272 |
+
"eval_wer": 0.4405900027314941,
|
1273 |
+
"step": 1600
|
1274 |
+
},
|
1275 |
+
{
|
1276 |
+
"epoch": 13.879310344827585,
|
1277 |
+
"grad_norm": 1.331200361251831,
|
1278 |
+
"learning_rate": 5.393939393939394e-05,
|
1279 |
+
"loss": 0.0648,
|
1280 |
+
"step": 1610
|
1281 |
+
},
|
1282 |
+
{
|
1283 |
+
"epoch": 13.96551724137931,
|
1284 |
+
"grad_norm": 2.28397536277771,
|
1285 |
+
"learning_rate": 5.3181818181818186e-05,
|
1286 |
+
"loss": 0.1121,
|
1287 |
+
"step": 1620
|
1288 |
+
},
|
1289 |
+
{
|
1290 |
+
"epoch": 14.051724137931034,
|
1291 |
+
"grad_norm": 0.9436893463134766,
|
1292 |
+
"learning_rate": 5.242424242424243e-05,
|
1293 |
+
"loss": 0.0725,
|
1294 |
+
"step": 1630
|
1295 |
+
},
|
1296 |
+
{
|
1297 |
+
"epoch": 14.137931034482758,
|
1298 |
+
"grad_norm": 1.6113288402557373,
|
1299 |
+
"learning_rate": 5.166666666666667e-05,
|
1300 |
+
"loss": 0.0691,
|
1301 |
+
"step": 1640
|
1302 |
+
},
|
1303 |
+
{
|
1304 |
+
"epoch": 14.224137931034482,
|
1305 |
+
"grad_norm": 2.479888439178467,
|
1306 |
+
"learning_rate": 5.090909090909091e-05,
|
1307 |
+
"loss": 0.0979,
|
1308 |
+
"step": 1650
|
1309 |
+
},
|
1310 |
+
{
|
1311 |
+
"epoch": 14.310344827586206,
|
1312 |
+
"grad_norm": 1.006616473197937,
|
1313 |
+
"learning_rate": 5.015151515151515e-05,
|
1314 |
+
"loss": 0.0909,
|
1315 |
+
"step": 1660
|
1316 |
+
},
|
1317 |
+
{
|
1318 |
+
"epoch": 14.39655172413793,
|
1319 |
+
"grad_norm": 1.4571704864501953,
|
1320 |
+
"learning_rate": 4.93939393939394e-05,
|
1321 |
+
"loss": 0.0761,
|
1322 |
+
"step": 1670
|
1323 |
+
},
|
1324 |
+
{
|
1325 |
+
"epoch": 14.482758620689655,
|
1326 |
+
"grad_norm": 1.5729875564575195,
|
1327 |
+
"learning_rate": 4.863636363636364e-05,
|
1328 |
+
"loss": 0.0862,
|
1329 |
+
"step": 1680
|
1330 |
+
},
|
1331 |
+
{
|
1332 |
+
"epoch": 14.568965517241379,
|
1333 |
+
"grad_norm": 1.2180376052856445,
|
1334 |
+
"learning_rate": 4.787878787878788e-05,
|
1335 |
+
"loss": 0.0646,
|
1336 |
+
"step": 1690
|
1337 |
+
},
|
1338 |
+
{
|
1339 |
+
"epoch": 14.655172413793103,
|
1340 |
+
"grad_norm": 1.7464072704315186,
|
1341 |
+
"learning_rate": 4.712121212121212e-05,
|
1342 |
+
"loss": 0.0741,
|
1343 |
+
"step": 1700
|
1344 |
+
},
|
1345 |
+
{
|
1346 |
+
"epoch": 14.655172413793103,
|
1347 |
+
"eval_loss": 0.4113341271877289,
|
1348 |
+
"eval_runtime": 40.2841,
|
1349 |
+
"eval_samples_per_second": 33.363,
|
1350 |
+
"eval_steps_per_second": 33.363,
|
1351 |
+
"eval_wer": 0.4309387234817445,
|
1352 |
+
"step": 1700
|
1353 |
+
},
|
1354 |
+
{
|
1355 |
+
"epoch": 14.741379310344827,
|
1356 |
+
"grad_norm": 0.8571386337280273,
|
1357 |
+
"learning_rate": 4.6439393939393944e-05,
|
1358 |
+
"loss": 0.1315,
|
1359 |
+
"step": 1710
|
1360 |
+
},
|
1361 |
+
{
|
1362 |
+
"epoch": 14.827586206896552,
|
1363 |
+
"grad_norm": 1.331377387046814,
|
1364 |
+
"learning_rate": 4.5681818181818186e-05,
|
1365 |
+
"loss": 0.0603,
|
1366 |
+
"step": 1720
|
1367 |
+
},
|
1368 |
+
{
|
1369 |
+
"epoch": 14.913793103448276,
|
1370 |
+
"grad_norm": 1.5398732423782349,
|
1371 |
+
"learning_rate": 4.492424242424242e-05,
|
1372 |
+
"loss": 0.0796,
|
1373 |
+
"step": 1730
|
1374 |
+
},
|
1375 |
+
{
|
1376 |
+
"epoch": 15.0,
|
1377 |
+
"grad_norm": 3.689671754837036,
|
1378 |
+
"learning_rate": 4.4166666666666665e-05,
|
1379 |
+
"loss": 0.085,
|
1380 |
+
"step": 1740
|
1381 |
+
},
|
1382 |
+
{
|
1383 |
+
"epoch": 15.086206896551724,
|
1384 |
+
"grad_norm": 1.132613182067871,
|
1385 |
+
"learning_rate": 4.340909090909091e-05,
|
1386 |
+
"loss": 0.0544,
|
1387 |
+
"step": 1750
|
1388 |
+
},
|
1389 |
+
{
|
1390 |
+
"epoch": 15.172413793103448,
|
1391 |
+
"grad_norm": 1.5951859951019287,
|
1392 |
+
"learning_rate": 4.265151515151515e-05,
|
1393 |
+
"loss": 0.0601,
|
1394 |
+
"step": 1760
|
1395 |
+
},
|
1396 |
+
{
|
1397 |
+
"epoch": 15.258620689655173,
|
1398 |
+
"grad_norm": 0.5179944634437561,
|
1399 |
+
"learning_rate": 4.189393939393939e-05,
|
1400 |
+
"loss": 0.097,
|
1401 |
+
"step": 1770
|
1402 |
+
},
|
1403 |
+
{
|
1404 |
+
"epoch": 15.344827586206897,
|
1405 |
+
"grad_norm": 0.9744370579719543,
|
1406 |
+
"learning_rate": 4.113636363636364e-05,
|
1407 |
+
"loss": 0.0596,
|
1408 |
+
"step": 1780
|
1409 |
+
},
|
1410 |
+
{
|
1411 |
+
"epoch": 15.431034482758621,
|
1412 |
+
"grad_norm": 1.8794275522232056,
|
1413 |
+
"learning_rate": 4.0378787878787885e-05,
|
1414 |
+
"loss": 0.0677,
|
1415 |
+
"step": 1790
|
1416 |
+
},
|
1417 |
+
{
|
1418 |
+
"epoch": 15.517241379310345,
|
1419 |
+
"grad_norm": 0.748386025428772,
|
1420 |
+
"learning_rate": 3.962121212121213e-05,
|
1421 |
+
"loss": 0.0896,
|
1422 |
+
"step": 1800
|
1423 |
+
},
|
1424 |
+
{
|
1425 |
+
"epoch": 15.517241379310345,
|
1426 |
+
"eval_loss": 0.43920788168907166,
|
1427 |
+
"eval_runtime": 40.1997,
|
1428 |
+
"eval_samples_per_second": 33.433,
|
1429 |
+
"eval_steps_per_second": 33.433,
|
1430 |
+
"eval_wer": 0.4307566238732587,
|
1431 |
+
"step": 1800
|
1432 |
+
},
|
1433 |
+
{
|
1434 |
+
"epoch": 15.60344827586207,
|
1435 |
+
"grad_norm": 0.9639837145805359,
|
1436 |
+
"learning_rate": 3.8863636363636364e-05,
|
1437 |
+
"loss": 0.0604,
|
1438 |
+
"step": 1810
|
1439 |
+
},
|
1440 |
+
{
|
1441 |
+
"epoch": 15.689655172413794,
|
1442 |
+
"grad_norm": 1.9640839099884033,
|
1443 |
+
"learning_rate": 3.810606060606061e-05,
|
1444 |
+
"loss": 0.0711,
|
1445 |
+
"step": 1820
|
1446 |
+
},
|
1447 |
+
{
|
1448 |
+
"epoch": 15.775862068965518,
|
1449 |
+
"grad_norm": 1.4438735246658325,
|
1450 |
+
"learning_rate": 3.734848484848485e-05,
|
1451 |
+
"loss": 0.0867,
|
1452 |
+
"step": 1830
|
1453 |
+
},
|
1454 |
+
{
|
1455 |
+
"epoch": 15.862068965517242,
|
1456 |
+
"grad_norm": 1.0062426328659058,
|
1457 |
+
"learning_rate": 3.659090909090909e-05,
|
1458 |
+
"loss": 0.0605,
|
1459 |
+
"step": 1840
|
1460 |
+
},
|
1461 |
+
{
|
1462 |
+
"epoch": 15.948275862068966,
|
1463 |
+
"grad_norm": 1.6331523656845093,
|
1464 |
+
"learning_rate": 3.5833333333333335e-05,
|
1465 |
+
"loss": 0.0662,
|
1466 |
+
"step": 1850
|
1467 |
+
},
|
1468 |
+
{
|
1469 |
+
"epoch": 16.03448275862069,
|
1470 |
+
"grad_norm": 0.8070217370986938,
|
1471 |
+
"learning_rate": 3.507575757575758e-05,
|
1472 |
+
"loss": 0.0765,
|
1473 |
+
"step": 1860
|
1474 |
+
},
|
1475 |
+
{
|
1476 |
+
"epoch": 16.120689655172413,
|
1477 |
+
"grad_norm": 1.4137670993804932,
|
1478 |
+
"learning_rate": 3.431818181818182e-05,
|
1479 |
+
"loss": 0.0537,
|
1480 |
+
"step": 1870
|
1481 |
+
},
|
1482 |
+
{
|
1483 |
+
"epoch": 16.20689655172414,
|
1484 |
+
"grad_norm": 1.5437769889831543,
|
1485 |
+
"learning_rate": 3.356060606060606e-05,
|
1486 |
+
"loss": 0.0684,
|
1487 |
+
"step": 1880
|
1488 |
+
},
|
1489 |
+
{
|
1490 |
+
"epoch": 16.29310344827586,
|
1491 |
+
"grad_norm": 0.90281081199646,
|
1492 |
+
"learning_rate": 3.2803030303030305e-05,
|
1493 |
+
"loss": 0.0744,
|
1494 |
+
"step": 1890
|
1495 |
+
},
|
1496 |
+
{
|
1497 |
+
"epoch": 16.379310344827587,
|
1498 |
+
"grad_norm": 1.139837622642517,
|
1499 |
+
"learning_rate": 3.204545454545455e-05,
|
1500 |
+
"loss": 0.0492,
|
1501 |
+
"step": 1900
|
1502 |
+
},
|
1503 |
+
{
|
1504 |
+
"epoch": 16.379310344827587,
|
1505 |
+
"eval_loss": 0.4201890528202057,
|
1506 |
+
"eval_runtime": 40.1502,
|
1507 |
+
"eval_samples_per_second": 33.474,
|
1508 |
+
"eval_steps_per_second": 33.474,
|
1509 |
+
"eval_wer": 0.4313029226987162,
|
1510 |
+
"step": 1900
|
1511 |
+
},
|
1512 |
+
{
|
1513 |
+
"epoch": 16.46551724137931,
|
1514 |
+
"grad_norm": 1.679457426071167,
|
1515 |
+
"learning_rate": 3.128787878787879e-05,
|
1516 |
+
"loss": 0.0652,
|
1517 |
+
"step": 1910
|
1518 |
+
},
|
1519 |
+
{
|
1520 |
+
"epoch": 16.551724137931036,
|
1521 |
+
"grad_norm": 0.6661111116409302,
|
1522 |
+
"learning_rate": 3.0530303030303034e-05,
|
1523 |
+
"loss": 0.0649,
|
1524 |
+
"step": 1920
|
1525 |
+
},
|
1526 |
+
{
|
1527 |
+
"epoch": 16.637931034482758,
|
1528 |
+
"grad_norm": 1.1774355173110962,
|
1529 |
+
"learning_rate": 2.9772727272727273e-05,
|
1530 |
+
"loss": 0.0469,
|
1531 |
+
"step": 1930
|
1532 |
+
},
|
1533 |
+
{
|
1534 |
+
"epoch": 16.724137931034484,
|
1535 |
+
"grad_norm": 1.783923864364624,
|
1536 |
+
"learning_rate": 2.901515151515152e-05,
|
1537 |
+
"loss": 0.0752,
|
1538 |
+
"step": 1940
|
1539 |
+
},
|
1540 |
+
{
|
1541 |
+
"epoch": 16.810344827586206,
|
1542 |
+
"grad_norm": 1.176321268081665,
|
1543 |
+
"learning_rate": 2.825757575757576e-05,
|
1544 |
+
"loss": 0.0519,
|
1545 |
+
"step": 1950
|
1546 |
+
},
|
1547 |
+
{
|
1548 |
+
"epoch": 16.896551724137932,
|
1549 |
+
"grad_norm": 1.3150608539581299,
|
1550 |
+
"learning_rate": 2.7500000000000004e-05,
|
1551 |
+
"loss": 0.0547,
|
1552 |
+
"step": 1960
|
1553 |
+
},
|
1554 |
+
{
|
1555 |
+
"epoch": 16.982758620689655,
|
1556 |
+
"grad_norm": 0.983769953250885,
|
1557 |
+
"learning_rate": 2.674242424242424e-05,
|
1558 |
+
"loss": 0.0799,
|
1559 |
+
"step": 1970
|
1560 |
+
},
|
1561 |
+
{
|
1562 |
+
"epoch": 17.06896551724138,
|
1563 |
+
"grad_norm": 0.996890127658844,
|
1564 |
+
"learning_rate": 2.5984848484848483e-05,
|
1565 |
+
"loss": 0.0577,
|
1566 |
+
"step": 1980
|
1567 |
+
},
|
1568 |
+
{
|
1569 |
+
"epoch": 17.155172413793103,
|
1570 |
+
"grad_norm": 2.3034253120422363,
|
1571 |
+
"learning_rate": 2.5227272727272726e-05,
|
1572 |
+
"loss": 0.0515,
|
1573 |
+
"step": 1990
|
1574 |
+
},
|
1575 |
+
{
|
1576 |
+
"epoch": 17.24137931034483,
|
1577 |
+
"grad_norm": 3.7528610229492188,
|
1578 |
+
"learning_rate": 2.4469696969696972e-05,
|
1579 |
+
"loss": 0.0759,
|
1580 |
+
"step": 2000
|
1581 |
+
},
|
1582 |
+
{
|
1583 |
+
"epoch": 17.24137931034483,
|
1584 |
+
"eval_loss": 0.43480169773101807,
|
1585 |
+
"eval_runtime": 40.017,
|
1586 |
+
"eval_samples_per_second": 33.586,
|
1587 |
+
"eval_steps_per_second": 33.586,
|
1588 |
+
"eval_wer": 0.4207411454065374,
|
1589 |
+
"step": 2000
|
1590 |
+
},
|
1591 |
+
{
|
1592 |
+
"epoch": 17.32758620689655,
|
1593 |
+
"grad_norm": 0.6646668314933777,
|
1594 |
+
"learning_rate": 2.3712121212121214e-05,
|
1595 |
+
"loss": 0.0419,
|
1596 |
+
"step": 2010
|
1597 |
+
},
|
1598 |
+
{
|
1599 |
+
"epoch": 17.413793103448278,
|
1600 |
+
"grad_norm": 1.3250740766525269,
|
1601 |
+
"learning_rate": 2.2954545454545457e-05,
|
1602 |
+
"loss": 0.0595,
|
1603 |
+
"step": 2020
|
1604 |
+
},
|
1605 |
+
{
|
1606 |
+
"epoch": 17.5,
|
1607 |
+
"grad_norm": 0.8094995021820068,
|
1608 |
+
"learning_rate": 2.21969696969697e-05,
|
1609 |
+
"loss": 0.0691,
|
1610 |
+
"step": 2030
|
1611 |
+
},
|
1612 |
+
{
|
1613 |
+
"epoch": 17.586206896551722,
|
1614 |
+
"grad_norm": 0.846946120262146,
|
1615 |
+
"learning_rate": 2.143939393939394e-05,
|
1616 |
+
"loss": 0.052,
|
1617 |
+
"step": 2040
|
1618 |
+
},
|
1619 |
+
{
|
1620 |
+
"epoch": 17.67241379310345,
|
1621 |
+
"grad_norm": 1.652417540550232,
|
1622 |
+
"learning_rate": 2.0681818181818182e-05,
|
1623 |
+
"loss": 0.0565,
|
1624 |
+
"step": 2050
|
1625 |
+
},
|
1626 |
+
{
|
1627 |
+
"epoch": 17.75862068965517,
|
1628 |
+
"grad_norm": 1.0080279111862183,
|
1629 |
+
"learning_rate": 1.9924242424242425e-05,
|
1630 |
+
"loss": 0.0745,
|
1631 |
+
"step": 2060
|
1632 |
+
},
|
1633 |
+
{
|
1634 |
+
"epoch": 17.844827586206897,
|
1635 |
+
"grad_norm": 0.7252691388130188,
|
1636 |
+
"learning_rate": 1.9166666666666667e-05,
|
1637 |
+
"loss": 0.0513,
|
1638 |
+
"step": 2070
|
1639 |
+
},
|
1640 |
+
{
|
1641 |
+
"epoch": 17.93103448275862,
|
1642 |
+
"grad_norm": 1.58548903465271,
|
1643 |
+
"learning_rate": 1.840909090909091e-05,
|
1644 |
+
"loss": 0.055,
|
1645 |
+
"step": 2080
|
1646 |
+
},
|
1647 |
+
{
|
1648 |
+
"epoch": 18.017241379310345,
|
1649 |
+
"grad_norm": 0.6634634733200073,
|
1650 |
+
"learning_rate": 1.7651515151515153e-05,
|
1651 |
+
"loss": 0.0658,
|
1652 |
+
"step": 2090
|
1653 |
+
},
|
1654 |
+
{
|
1655 |
+
"epoch": 18.103448275862068,
|
1656 |
+
"grad_norm": 1.1495524644851685,
|
1657 |
+
"learning_rate": 1.6893939393939395e-05,
|
1658 |
+
"loss": 0.0406,
|
1659 |
+
"step": 2100
|
1660 |
+
},
|
1661 |
+
{
|
1662 |
+
"epoch": 18.103448275862068,
|
1663 |
+
"eval_loss": 0.44191813468933105,
|
1664 |
+
"eval_runtime": 40.0967,
|
1665 |
+
"eval_samples_per_second": 33.519,
|
1666 |
+
"eval_steps_per_second": 33.519,
|
1667 |
+
"eval_wer": 0.42046799599380863,
|
1668 |
+
"step": 2100
|
1669 |
+
},
|
1670 |
+
{
|
1671 |
+
"epoch": 18.189655172413794,
|
1672 |
+
"grad_norm": 0.9788354635238647,
|
1673 |
+
"learning_rate": 1.6136363636363638e-05,
|
1674 |
+
"loss": 0.0381,
|
1675 |
+
"step": 2110
|
1676 |
+
},
|
1677 |
+
{
|
1678 |
+
"epoch": 18.275862068965516,
|
1679 |
+
"grad_norm": 1.093633770942688,
|
1680 |
+
"learning_rate": 1.5378787878787877e-05,
|
1681 |
+
"loss": 0.071,
|
1682 |
+
"step": 2120
|
1683 |
+
},
|
1684 |
+
{
|
1685 |
+
"epoch": 18.362068965517242,
|
1686 |
+
"grad_norm": 0.7164376974105835,
|
1687 |
+
"learning_rate": 1.4621212121212122e-05,
|
1688 |
+
"loss": 0.0439,
|
1689 |
+
"step": 2130
|
1690 |
+
},
|
1691 |
+
{
|
1692 |
+
"epoch": 18.448275862068964,
|
1693 |
+
"grad_norm": 0.9887032508850098,
|
1694 |
+
"learning_rate": 1.3863636363636364e-05,
|
1695 |
+
"loss": 0.0481,
|
1696 |
+
"step": 2140
|
1697 |
+
},
|
1698 |
+
{
|
1699 |
+
"epoch": 18.53448275862069,
|
1700 |
+
"grad_norm": 0.45052286982536316,
|
1701 |
+
"learning_rate": 1.3106060606060607e-05,
|
1702 |
+
"loss": 0.0571,
|
1703 |
+
"step": 2150
|
1704 |
+
},
|
1705 |
+
{
|
1706 |
+
"epoch": 18.620689655172413,
|
1707 |
+
"grad_norm": 1.167181134223938,
|
1708 |
+
"learning_rate": 1.234848484848485e-05,
|
1709 |
+
"loss": 0.0452,
|
1710 |
+
"step": 2160
|
1711 |
+
},
|
1712 |
+
{
|
1713 |
+
"epoch": 18.70689655172414,
|
1714 |
+
"grad_norm": 1.378661870956421,
|
1715 |
+
"learning_rate": 1.159090909090909e-05,
|
1716 |
+
"loss": 0.0643,
|
1717 |
+
"step": 2170
|
1718 |
+
},
|
1719 |
+
{
|
1720 |
+
"epoch": 18.79310344827586,
|
1721 |
+
"grad_norm": 0.854932963848114,
|
1722 |
+
"learning_rate": 1.0833333333333334e-05,
|
1723 |
+
"loss": 0.0587,
|
1724 |
+
"step": 2180
|
1725 |
+
},
|
1726 |
+
{
|
1727 |
+
"epoch": 18.879310344827587,
|
1728 |
+
"grad_norm": 0.8007526397705078,
|
1729 |
+
"learning_rate": 1.0075757575757576e-05,
|
1730 |
+
"loss": 0.0395,
|
1731 |
+
"step": 2190
|
1732 |
+
},
|
1733 |
+
{
|
1734 |
+
"epoch": 18.96551724137931,
|
1735 |
+
"grad_norm": 3.317830801010132,
|
1736 |
+
"learning_rate": 9.318181818181819e-06,
|
1737 |
+
"loss": 0.074,
|
1738 |
+
"step": 2200
|
1739 |
+
},
|
1740 |
+
{
|
1741 |
+
"epoch": 18.96551724137931,
|
1742 |
+
"eval_loss": 0.43061742186546326,
|
1743 |
+
"eval_runtime": 40.0034,
|
1744 |
+
"eval_samples_per_second": 33.597,
|
1745 |
+
"eval_steps_per_second": 33.597,
|
1746 |
+
"eval_wer": 0.420012746972594,
|
1747 |
+
"step": 2200
|
1748 |
+
},
|
1749 |
+
{
|
1750 |
+
"epoch": 19.051724137931036,
|
1751 |
+
"grad_norm": 0.7710875272750854,
|
1752 |
+
"learning_rate": 8.56060606060606e-06,
|
1753 |
+
"loss": 0.046,
|
1754 |
+
"step": 2210
|
1755 |
+
},
|
1756 |
+
{
|
1757 |
+
"epoch": 19.137931034482758,
|
1758 |
+
"grad_norm": 0.5200530886650085,
|
1759 |
+
"learning_rate": 7.803030303030304e-06,
|
1760 |
+
"loss": 0.0394,
|
1761 |
+
"step": 2220
|
1762 |
+
},
|
1763 |
+
{
|
1764 |
+
"epoch": 19.224137931034484,
|
1765 |
+
"grad_norm": 1.3544327020645142,
|
1766 |
+
"learning_rate": 7.045454545454545e-06,
|
1767 |
+
"loss": 0.0582,
|
1768 |
+
"step": 2230
|
1769 |
+
},
|
1770 |
+
{
|
1771 |
+
"epoch": 19.310344827586206,
|
1772 |
+
"grad_norm": 0.8653574585914612,
|
1773 |
+
"learning_rate": 6.287878787878789e-06,
|
1774 |
+
"loss": 0.0606,
|
1775 |
+
"step": 2240
|
1776 |
+
},
|
1777 |
+
{
|
1778 |
+
"epoch": 19.396551724137932,
|
1779 |
+
"grad_norm": 1.5852700471878052,
|
1780 |
+
"learning_rate": 5.530303030303031e-06,
|
1781 |
+
"loss": 0.0367,
|
1782 |
+
"step": 2250
|
1783 |
+
},
|
1784 |
+
{
|
1785 |
+
"epoch": 19.482758620689655,
|
1786 |
+
"grad_norm": 2.2167246341705322,
|
1787 |
+
"learning_rate": 4.772727272727273e-06,
|
1788 |
+
"loss": 0.0782,
|
1789 |
+
"step": 2260
|
1790 |
+
},
|
1791 |
+
{
|
1792 |
+
"epoch": 19.56896551724138,
|
1793 |
+
"grad_norm": 0.5891330242156982,
|
1794 |
+
"learning_rate": 4.015151515151515e-06,
|
1795 |
+
"loss": 0.0416,
|
1796 |
+
"step": 2270
|
1797 |
+
},
|
1798 |
+
{
|
1799 |
+
"epoch": 19.655172413793103,
|
1800 |
+
"grad_norm": 1.1137330532073975,
|
1801 |
+
"learning_rate": 3.257575757575758e-06,
|
1802 |
+
"loss": 0.0515,
|
1803 |
+
"step": 2280
|
1804 |
+
},
|
1805 |
+
{
|
1806 |
+
"epoch": 19.74137931034483,
|
1807 |
+
"grad_norm": 0.8132285475730896,
|
1808 |
+
"learning_rate": 2.5e-06,
|
1809 |
+
"loss": 0.0512,
|
1810 |
+
"step": 2290
|
1811 |
+
},
|
1812 |
+
{
|
1813 |
+
"epoch": 19.82758620689655,
|
1814 |
+
"grad_norm": 0.7994781136512756,
|
1815 |
+
"learning_rate": 1.7424242424242427e-06,
|
1816 |
+
"loss": 0.0378,
|
1817 |
+
"step": 2300
|
1818 |
+
},
|
1819 |
+
{
|
1820 |
+
"epoch": 19.82758620689655,
|
1821 |
+
"eval_loss": 0.4273350238800049,
|
1822 |
+
"eval_runtime": 40.0934,
|
1823 |
+
"eval_samples_per_second": 33.522,
|
1824 |
+
"eval_steps_per_second": 33.522,
|
1825 |
+
"eval_wer": 0.41728125284530637,
|
1826 |
+
"step": 2300
|
1827 |
+
},
|
1828 |
+
{
|
1829 |
+
"epoch": 19.913793103448278,
|
1830 |
+
"grad_norm": 0.9775754809379578,
|
1831 |
+
"learning_rate": 9.848484848484847e-07,
|
1832 |
+
"loss": 0.0489,
|
1833 |
+
"step": 2310
|
1834 |
+
},
|
1835 |
+
{
|
1836 |
+
"epoch": 20.0,
|
1837 |
+
"grad_norm": 0.8857516050338745,
|
1838 |
+
"learning_rate": 2.2727272727272726e-07,
|
1839 |
+
"loss": 0.0554,
|
1840 |
+
"step": 2320
|
1841 |
+
},
|
1842 |
+
{
|
1843 |
+
"epoch": 20.0,
|
1844 |
+
"step": 2320,
|
1845 |
+
"total_flos": 2.1476719263248095e+18,
|
1846 |
+
"train_loss": 0.8618391515622879,
|
1847 |
+
"train_runtime": 3159.4128,
|
1848 |
+
"train_samples_per_second": 23.397,
|
1849 |
+
"train_steps_per_second": 0.734
|
1850 |
+
}
|
1851 |
+
],
|
1852 |
+
"logging_steps": 10,
|
1853 |
+
"max_steps": 2320,
|
1854 |
+
"num_input_tokens_seen": 0,
|
1855 |
+
"num_train_epochs": 20,
|
1856 |
+
"save_steps": 400,
|
1857 |
+
"stateful_callbacks": {
|
1858 |
+
"TrainerControl": {
|
1859 |
+
"args": {
|
1860 |
+
"should_epoch_stop": false,
|
1861 |
+
"should_evaluate": false,
|
1862 |
+
"should_log": false,
|
1863 |
+
"should_save": true,
|
1864 |
+
"should_training_stop": false
|
1865 |
+
},
|
1866 |
+
"attributes": {}
|
1867 |
+
}
|
1868 |
+
},
|
1869 |
+
"total_flos": 2.1476719263248095e+18,
|
1870 |
+
"train_batch_size": 32,
|
1871 |
+
"trial_name": null,
|
1872 |
+
"trial_params": null
|
1873 |
+
}
|